We have developed a mouse model system, using hypomorphic alleles of the major DNA methyltransferase Dnmt1, in which we can assess the role of DNA methylation in oncogenesis and investigate various mechanisms by which DNA methylation and/or Dnmt1 may affect the cancer process. We have recently shown that combinations of Dnmt1 hypomorphic alleles can achieve complete genetic suppression of ApcMin/+ induced polyp formation, suggesting that sufficient levels of Dnmt1 expression are required for intestinal polyp development. This is further supported by our observation that intestinal tumorigenesis in mismatch-repair deficient Mih1-/- mice is also suppressed by low levels of Dnmt1. However, lymphomagenesis is increased in these same mice, suggesting that modulation of Dnmt1 levels can have opposing effects on oncogenesis in vivo. The molecular basis for this strong effect of Dnmt1 levels on various models of oncogenesis is not understood. We have recently shown in a tissue-culture model system that Dnmt1 deficiency causes a reduction in methylation-dependent genetic events, such as methylcytosine deamination, and a reduction in methylation-dependent epigenetic events, such as transcriptional silencing by promoter CpG island hypermethylation. In this competing continuation, we propose to 1) expand our analysis of the effects of in vivo modulation of DNA methylation on cancer model systems and 2) analyze the influence of DNA methylation and/or methyltransferases on mutation frequencies in vivo, and 3) analyze sequence features affecting de novo methylation in ES cells. In summary, we propose to continue to investigate the causal effects of DNA methylation in cancer and to analyze both genetic and epigenetic mechanisms by which DNA methylation and/or DNA methyltransferases could affect oncogenesis.